Cowbell today revealed it will be applying generative artificial intelligence to enable underwriters to process IT and cybersecurity insurance policies more efficiently and accurately.
Rajeev Gupta, chief product officer for Cowbell, said Cowbell Co-Pilot is making use of a large language model (LLM) developed by OpenAI within PrimeTech, an insurance offering that combines cyber insurance with a technology errors and omission policy.
The goal is to both reduce the number of underwriters needed to craft polices while improving overall accuracy, said Gupta. Most insurance policies today consist of a mix of insurance and technology jargon that in the absence of a summarization capability provided by AI is challenging for underwriters to decipher, he noted.
Cowbell Co-Pilot can also retrieve potentially relevant information from contracts and other sources to augment the underwriting process, leading to better risk assessments. In addition, there are often relationships between multiple polices that organizations have in place that will soon be able to surface more easily using generative AI tools, said Gupta. “They tend to bleed into each other,” he says.
According to Cowbell, Prime Tech with Co-Pilot will help underwriters reduce, on average, 40% of contract review times.
Prime Tech is available now to businesses with up to $250M in revenue in the U.S, with a limit of $5 million in coverage. Longer term, Cowbell expect to make similar AI capabilities available to the organizations that pay the premiums for this and other types of insurance, added Gupta.
Cowbell Co-Pilot doesn’t replace the need for underwriters, but it does promise to improve productivity to the point where either more policies can be reviewed faster or the total number of underwriters that might need to be required is less. The one thing that is certain is the amount of time that needs to be allocated to reviewing insurance policies will be substantially less.
Generative AI will undoubtedly soon be applied to underwriting for just about every type of insurance there is. Historically, the insurance industry has tended to be a laggard when it comes to adopting emerging technologies, but given the volume of data that insurance carriers need to process, it’s only a matter of time before AI is applied in a way that dramatically improves productivity. The issue is making sure that auditable protocols are in place that guarantee sensitive data that is included in an insurance policy is not inadvertently used to train an LLM that is publicly accessible to millions of end users. Otherwise, it’s only matter of time before that data manifests itself in a way that violates one or more regulations. In fact, until processes that safeguard data are in place it’s probable many carriers will proceed with caution before making extensive use of generative AI.
In the meantime, however, insurance providers should be, at the very least, assessing which processes might soon be automated using generative AI. The simple truth is many underwriters might already be taking advantage of generative AI without any supervision at all, so the sooner the right guardrails are put in place the better off everyone affected is likely to be.
